The Problem: The same housing data can tell very different stories depending on design choices. Our team made honest and misleading versions of the same dataset to show how much power a chart designer has.

 

My Role: I built an honest chart comparing median prices by property type and bedroom count.

 

Process

1. Started with a stacked bar chart of average prices, but 7-bedroom condos (around $200M on average) drowned out everything else

2. Switched to medians and split condos, houses, and townhouses into side-by-side small multiples

3. Added a log scale so $500K studios and $30M penthouses could both be read, then tested it with four people outside class

 

Result: Every tester said the log version let them compare across property types, while the original was basically unreadable. Based on their feedback, I added a clear "(log)" label and filtered out listings over 8 bedrooms.

 

What I Learned: Averages can mislead with outlier-heavy data like real estate, and small choices like scale and bin size change what people believe.